A Term Appeared 26 Times in HTML, but Never in the Visible Body
Published
This article was first published in Korean by Intrix Lab on August 19, 2026.
A term found 26 times across the full HTML did not appear once in the visible body. On one law firm's homepage, we counted 26 occurrences of a professional organization's name in the source and 0 in the on-screen body. These were measurements taken on 2026-08-19. Counting terms in schema and metadata as if they also appeared in the body can lead to a false conclusion that information is available through a path that reads only the body.
Where were the 26 occurrences?
First, we counted occurrences of the same professional organization's name across the full HTML source of the live homepage. The count was 26. We then counted it separately in only the body content visible on screen. The result was 0.
All 26 occurrences were in JSON-LD, the RSC payload, and the meta description. None appeared in the sentences people read on screen. Searching the full HTML confirmed that the term existed, but did not distinguish which layer contained it.
After the change, the professional organization's name was included in the visible body. A new measurement confirmed 1 occurrence in the homepage's visible body.
How does this finding change the diagnostic approach?
Search locations need to be separated. Full HTML, structured data, metadata, and the visible body belong to the same document, but they are not the same input. The presence of a verification term in schema alone does not establish that it also appears in the body.
The lesson from this case is not a prescription to add more of a particular term. It is that terms in schema and terms in the on-screen body need to be examined separately. When the two differ, the first step is to identify which layer is missing the signal.
How far can these results be trusted?
The population is a single site belonging to one law firm, and the period is a single measurement date, 2026-08-19. There is no guarantee that other industries or rendering structures will show a split like 26 occurrences versus 0. These figures cannot be generalized into a rate for legal websites as a whole.
The identification of visible body content also has limitations. Elements hidden with CSS were not filtered out, so the visible body as defined here does not exactly match the browser display.
This observation also does not support a causal conclusion that structured data is meaningless. The finding extends only to the difference between the full HTML count and the body count on one site. Intrix does not guarantee rankings, citations, or visibility.
Where should other sites start?
Counting the same term separately in the full HTML and the visible body provides a starting point. To check which layers contain signals in your site's current rendering and structured data, get measured results through the free SEO and GEO audit.
Frequently asked questions (FAQ)
Q. Can search engines not read information in JSON-LD?
This measurement did not test whether they could read it. It separately counted terms in JSON-LD and in the visible body. It cannot be interpreted as evidence about structured data processing or ranking effects.
Q. Should information included in schema also appear in the body?
Yes, if there is a path that reads only the body. In this case, the professional organization's name appeared only in schema and metadata, not in the on-screen body. After it was added to the body, 1 occurrence was confirmed in the homepage's visible body.